On 16 August 1972, Stefano Mariottini noticed an arm emerging from the seabed near Riace Marina, along Calabria’s Ionian coast in southern Italy. The arm belonged to a bronze statue almost two metres tall, and another figure lay close to it. Divers brought the sculptures ashore on 21 and 22 August. Their beards, faces and carefully modelled veins were still visible, along with an extraordinary amount of the workmanship that Greek sculptors had put into them. The two figures soon became known as the Riace Bronzes, or Riace Warriors.

Both statues date to the fifth century BCE, when Greek sculptors were finding increasingly naturalistic ways to represent the human body. Very few original Greek bronzes of comparable size survive from this period. Known simply as Bronze A and Bronze B, the figures once held weapons and were almost certainly displayed as armed men. Without their inscriptions, equipment or original setting, however, we cannot say securely whom they represent, who made them or where they first stood.

Their survival needs some explanation. Bronze was valuable enough to melt down for reuse, while seawater attacks copper alloys and can eventually destroy exposed surfaces. The Riace figures escaped the melting pot, but the sea did change them. Weapons and other fittings disappeared, corrosion altered the metal, and extensive conservation followed their discovery. Investigators have since studied the bronze, the material left inside the hollow castings and the deposits that once covered them to understand both their ancient manufacture and their long underwater history.

What Was Actually Found at Riace?

Mariottini’s report, filed the following day, placed the two nude male statues around 300 metres from shore in water roughly ten metres deep. The museum now gives their recovery depth as about eight metres. Those slightly different measurements refer to accounts of the same discovery, rather than different findspots. The familiar story that Mariottini initially thought the exposed arm belonged to a corpse is harder to establish than the details recorded in his contemporary written report.

With no inscriptions to provide names, researchers labelled the statues A and B. Bronze A, usually described as the younger warrior, measures approximately 1.98 metres; Bronze B stands around 1.97 metres. Both are taller than most men, and both balance their weight on the right leg, leaving the left more relaxed and shifting the hips and shoulders accordingly. The likeness in posture also draws attention to their differences. A has densely worked hair and a richly detailed beard; B’s head was shaped to accommodate headgear, and the treatment of his face and body is noticeably different.

The National Archaeological Museum of Reggio Calabria, which houses the bronzes, identifies them as armed warriors. Spears, shields and helmets have vanished, so their original appearance has to be inferred from the positions of their hands and arms and the traces of attachments. The lost equipment also makes identification more difficult. A nude figure now resembling an athlete might have been recognizable as a particular warrior, founder or legendary hero when seen with his weapons and the other figures or objects in his original monument.

Why So Few Greek Bronze Statues Survive

Greek artists made large statues in both marble and bronze, but the materials faced different dangers after their monuments fell out of use. Marble might be broken up for building or burned to make lime. Bronze could be melted and poured into entirely different objects. Because copper and tin had continuing economic value, an unwanted statue was also a source of metal for tools, fittings or another sculpture. Very little had to remain of the first work once it entered that cycle of reuse.

Greek cities once displayed monumental bronzes in large numbers, although very few survive today. In Natural History 34.36, Pliny the Elder says that the Roman official Mucianus counted 3,000 statues at Rhodes and reports similarly impressive numbers for Athens, Olympia and Delphi. His figures cannot be treated as a verified modern census, but they suggest how crowded sanctuaries and public places could be with bronze images. Of the original full-size Greek works, comparatively few remain. For many famous sculptures, Roman marble copies, smaller representations and descriptions are now the principal evidence.

The Metropolitan Museum of Art’s account of Greek bronze casting explains both the sophistication of bronze production and why the material was so readily recycled. Another useful comparison is the Boxer at Rest, an original Greek bronze kept from destruction by its unusual concealment. Burial and shipwrecks sometimes saved statues for precisely the same reason: once hidden, they were no longer available to people looking for metal to reuse.

After the Riace figures entered the sea, no one could readily take them apart for their copper and tin. The sea kept their metal beyond the reach of recyclers, although their earlier movements remain unknown. Their recovery place does not identify their former monument, the date when they were submerged or the circumstances of their loss. In particular, two statues found underwater do not, by themselves, prove that a ship sank with them.

Advertisements

What Seawater Does to Bronze

Bronze is made chiefly from copper alloyed with tin. Sculptors could cast it into demanding forms and refine the cooled surface with tools and abrasives, but exposure to seawater starts a different process. Oxygen dissolved in water, chloride ions and other compounds react with the copper. Depending on conditions, the metal may acquire red cuprite, dark copper sulfides or green copper chlorides. These products can build up in layers, while attack continues through cracks and other exposed areas.

A corrosion layer sometimes slows further contact with seawater, although it can also be porous or unstable. Bringing the object into air may trigger changes that were less active underwater. A dark coating is no guarantee that the metal beneath it is safe, and a green patina is not necessarily protective. Conservators therefore examine the composition, attachment and behaviour of deposits, especially where chloride remains. A layer that preserves traces of the ancient casting may also conceal active corrosion.

On the Riace warriors, substantial areas of the original modelling survived beneath marine deposits. Early analysis found a dark layer that had protected the metal in certain places. A Smithsonian Museum Conservation Institute bibliographic record discussing surface spectroscopy notes this beneficial black patina, together with chlorine and manganese connected to the marine environment. Its effect should not be imagined as a single watertight shield covering the statues: protection varied from one part of the surface to another.

The corrosion also points to changes in the conditions surrounding the statues. Copper oxides tend to form where oxygen is present, while copper sulfides may develop in oxygen-poor conditions, sometimes involving microbial processes. Chloride-rich deposits indicate further reactions with the marine environment. Because these materials are distributed unevenly, one patch of patina may reflect a different combination of sediment, water chemistry, biological growth and exposed bronze from another patch only a short distance away.

Six bronze corrosion samples showing green, blue, and brown patina layers from underwater archaeological helmets
Representative surface corrosion on archaeological bronze helmets showing different patina groups. Source: Archaeological conservation research.

The danger can increase once marine bronze is excavated. Chlorides trapped within copper alloys may take part in further reactions when air and moisture reach them, producing the damaging cycle commonly called bronze disease. The Getty’s technical publication on chloride-contaminated ancient bronzes discusses the difficulty of stopping that deterioration while retaining meaningful patina. Cleaning a surface indiscriminately could destroy evidence of where the object lay and how its metal changed over time.

Did the Seabed Protect the Statues for Two Thousand Years?

When the bronzes were discovered, they lay on sand in relatively shallow water. Sediment can partly shelter metal from abrasion by currents and from oxygen circulating in the water. Burial may also create a different chemical environment around one side of an object. Marine organisms sometimes leave calcareous deposits that help limit physical wear. All these processes offer possible explanations for why fine anatomical details remained visible beneath the accumulated material.

These conditions tell us little about whether the statues remained in the same place throughout their immersion. Shallow coastal water changes with seasons and storms, and shifting sand can expose or cover an object repeatedly. Even at a depth of about eight metres, conditions are neither uniformly calm nor thermally fixed. Without stronger archaeological and sedimentary evidence, it is impossible to establish a precise year-round temperature or claim that the bronzes remained in the same position throughout their immersion.

A 2026 study by Rosolino Cirrincione and fourteen co-authors proposes a more complicated history. Comparing published corrosion research, geological observations and clay samples from the Syracuse area of Sicily, the researchers examined the order in which different deposits may have formed. They interpret cuprite, a dark copper-sulfide layer, calcareous material and later chloride-rich deposits as signs of a long stay in deeper and less turbulent water before the bronzes reached Riace. Their proposed earlier depth is approximately 70 to 100 metres.

Cirrincione and his colleagues offer a reconstruction rather than a documented itinerary. The researchers also connect their chemical interpretation with a disputed claim that the statues were initially found near Syracuse and transferred to Riace in modern times. Their argument combines laboratory comparisons, marine biological evidence and historical proposals, but it does not identify an excavated shipwreck or document a journey from one findspot to the other. Both the estimated depth and the suggested Sicilian connection require further testing.

The study challenges the assumption that a sheltered stretch of shallow water accounts for the condition of the bronzes, although its reconstruction remains disputed. Deposits made by corrosion and marine organisms can record earlier surroundings that are no longer visible at the recovery site. The length of their immersion, the places in which they lay and the date when they first entered the sea cannot yet be reconstructed with precision from that evidence.

Turquoise Ionian Sea waters with rocky islands visible under clear Mediterranean sky near Lefkada Greece
Ionian Sea coastline photographed from Lefkada, Greece, 2014. Source: Wikimedia Commons
Advertisements

How the Warriors Were Made

The figures also preserve unusually detailed evidence of fifth-century BCE Greek bronze casting. The sculptors used lost-wax casting. In this method a wax model occupies the space eventually filled with molten bronze, between a prepared heat-resistant core and an outer mould. Heating removes the wax; liquid metal then enters the cavity. Once the casting has cooled, the mould is removed and the bronze can be cut, smoothed and finished. It is a demanding technique, but it permits intricate shapes that would be difficult to produce otherwise.

A statue nearly two metres tall was not cast as a solid mass of bronze. Both Riace figures have hollow bodies made from separately cast pieces that were assembled and finished together. The joins were worked carefully enough to make the surface appear continuous. Controlling the thickness of the walls and fitting limbs and other sections together demanded considerable skill. Hollow casting also used less metal than a solid figure would have required, although a finished monumental bronze remained a costly undertaking.

The Met’s technical account of bronze statuary describes the cold-working that followed casting, when chisels and other tools sharpened features such as hair, lips and muscles. Sculptors could also insert details made in different materials. On the Riace warriors, the beard curls, tendons, veins and subtle differences between supporting and relaxed muscles still show how the cast form was refined by hand. These anatomical effects were conscious decisions by the makers, preserved through the centuries rather than produced by corrosion.

Close-up faces of Riace Bronzes showing bearded warriors with green bronze patina and intricate hair detail
Heads of Riace Bronze Statue A and Statue B showing anatomical precision, 5th century BCE. Source: Finestresullarte.info

Each warrior has his own character of modelling. A’s animated beard and hair sit above a body that seems ready to move, even while the figure stands still. B carries his head and torso differently, and his helmet once changed the outline of his head. Such distinctions suggest that each was intended to be recognized as an individual within a possible larger group. They do not, in the absence of the original monument, tell us which historical or mythical episode the two men represented.

Eyes, Teeth and the Colour of Greek Bronze

Their appearance in antiquity would also have been much more varied in colour than the green or dark bronze familiar from weathered sculpture. The museum records calcite and glass paste in the eyes, with copper and silver used for other anatomical details. Bronze A still has silver teeth visible inside his partly open mouth. Copper inlays give the lips and nipples another tone, while separate pale and dark elements construct the eyes. Marks left by attachments and finishing treatments add further evidence of their original appearance.

The surviving inlays help explain the lifelike appearance of the faces. The bronze itself was one colour among several, supplemented by stone or glass in the eyes and differently coloured metals. Its ancient finish would have looked unlike the corroded surface visible after centuries underwater. Spears, shields and helmets added still more to the composition. Reconstructions can suggest how the complete figures might have appeared, but lost fittings and particular colour treatments are not always recoverable from the evidence.

Vinzenz Brinkmann and Ulrike Koch-Brinkmann have tested those possibilities in full-scale experimental copies, using detailed examination and three-dimensional recordings of the originals. Their published account of the project describes the effect of reproducing eyes, eyelashes, lips, teeth and nipples with materials chosen to resemble those of antiquity. The Metropolitan Museum’s record of the reconstructed Warrior A makes clear that proposed weapons and identification remain interpretive choices. These replicas test ancient techniques and visual possibilities; they cannot restore a uniquely certain original appearance.

The surviving traces of colour on stone sculpture, discussed in Greek statues and their lost colours, show that ancient Greek statuary was rarely as monochrome as later tastes assumed. The Riace bronzes preserve that evidence directly: several of their contrasting metals and eye materials still occupy the positions where the sculptors placed them.

Advertisements

The Clay Inside the Statues: Evidence for Their Age and Origin

Inside the hollow figures, conservators encountered large amounts of casting-core material, about 72 kilograms from Bronze A and 56 kilograms from Bronze B. Rather than discarding this as debris, researchers examined its clay, sand, mineral particles and organic inclusions. Such mixtures can preserve evidence of how a workshop prepared its moulds and where some of the ingredients may have been gathered.

A 1998 study by G. Lombardi and Massimo Vidale compared the cores geochemically. Differences among the samples opened the possibility of separate workshops or raw materials from more than one location. Geological resemblance cannot, however, identify a foundry by itself. Similar sediments occur in different areas, and a sculptor could cast, modify or assemble different parts of the same figure at different times. These comparisons cannot establish where the bronzes were made.

The organic material in the cores provides another line of evidence. Charcoal, plant matter and animal hair were recovered and dated using accelerator mass spectrometry, a radiocarbon technique. A study by Gianluca Quarta, Lucio Calcagnile and Massimo Vidale combined several measurements and gave calibrated ranges of 512–398 BCE for Bronze A and 511–400 BCE for Bronze B, both at 95.4% probability. Those broad ranges fit production during the fifth century BCE.

Radiocarbon results cannot resolve the much narrower question of whether A preceded B by about thirty years. The calibration curve is fairly flat across part of the relevant period, limiting the precision available. Art historians have often placed Bronze A around 470–460 BCE and B around 440–430 BCE on stylistic grounds. These are informed comparisons, not dates demonstrated by the laboratory. Both bronzes can be assigned broadly to the fifth century without establishing whether they were cast as a pair or separated by decades.

The core analyses also found marked differences between A and B, and between Bronze B’s right arm and the rest of that statue. A separately prepared arm could indicate ancient repair or modification, and the result points to an additional stage of manufacture. The radiocarbon studies did not independently date the arm, however. Nor can a secure century for that alteration be obtained simply by citing the percentage of lead in its metal.

Technical diagram showing lost-wax bronze casting process with clay core wax model and molten metal pouring stages
Bronze casting by the lost-wax method used for ancient Greek statues. Source: Wikimedia Commons

The workshop remains unidentified. The museum discusses possible Greek origins in regions such as Attica and Argolis. The 2026 geological study, by contrast, finds potential connections between some core material and the area of Sybaris in southern Italy, and compares material related to assembly with clay near Syracuse. The differences might reflect the separate places and stages involved in manufacture and installation. Neither the names of the sculptors nor the location of the original monument have been securely established.

Were They Lost in a Shipwreck?

Two monumental statues might readily have entered the sea while being carried on a ship. Sculptures travelled around the Mediterranean, and Roman patrons in particular collected Greek works in large numbers. A storm might have wrecked a vessel, or the crew might have jettisoned heavy cargo to save it. Either possibility could explain how the figures became inaccessible to metal recyclers, although it would also have exposed them to the destructive effects of salt water.

Investigators have still not demonstrated a shipwreck at Riace. Finds made after 1972 included lead rings, pottery fragments and some timber, but none has been securely associated with a vessel carrying both bronzes. The Italian Enciclopedia dell’Arte Antica’s discussion of the discovery warned against treating these scattered objects as evidence for one identifiable cargo. According to the 2026 geological study, subsequent underwater searches, including investigations in 2024, likewise failed to locate a wreck connected to the statues.

We consequently cannot tell whether both statues fell into the sea on the same Roman-era voyage, entered the water earlier, or passed through several different locations before their recovery. Transport to a Roman buyer remains plausible because the Roman appetite for Greek sculpture is well documented. No shipping record, however, mentions these particular works. The Riace discovery is firm evidence of where the figures were recovered in the twentieth century, not a map of their earlier movements.

Whatever brought the bronzes into the sea also placed them beyond the easy reach of metal recyclers. Hidden underwater, they could not easily be melted down, even as the surrounding sea slowly altered them. Their preservation is exceptional among large Greek bronzes. Yet the physical evidence of the metal and deposits is more reliable than an attractive shipwreck story for which the ship has not been found.

Restoring the Riace Bronzes After 1972

Bringing the bronzes ashore began another chapter of their history. Sand, encrustations and corrosion covered parts of their bodies, while material from the original casting remained inside. Following initial treatment in Reggio Calabria, a major conservation programme was carried out in Florence from 1975 to 1980. Specialists removed obstructive deposits and investigated construction details as they worked. The statues’ subsequent exhibition in Florence brought them to an international audience.

During the early 1990s, work inside the hollow figures became more systematic. Conservators studied and removed surviving core material, creating the samples subsequently used for radiocarbon and geological analysis. Access to the interiors also revealed construction details that had been difficult to inspect immediately after recovery. Those examinations helped separate ancient joints and modifications from changes caused by prolonged immersion or modern handling.

Conservators cannot always remove dangerous deposits without losing information. A deposit may protect the ancient surface while holding salts that threaten its future stability; cleaning may disclose a sculptor’s marks while erasing evidence of the object’s underwater environment. Desalination and stabilization are therefore chosen according to the needs of each object. Technical publications describe many possible procedures, but without the Riace treatment records it would be wrong to assume that every general conservation method was applied to these particular statues.

Conservator in white coat examining ancient bronze statue during restoration treatment in museum laboratory
Conservation specialist treating ancient bronze sculpture in restoration laboratory. Source: Liebieghaus Skulpturensammlung.

Another extensive conservation programme ran from 2010 to 2013. The bronzes returned to the museum on 21 December 2013. The Italian Ministry of Culture’s announcement describes their display room’s air-monitoring and filtration equipment and the specially constructed anti-seismic bases of Carrara marble. Environmental control helps limit corrosion risks, while the bases are designed to protect two tall, hollow figures in the event of an earthquake.

Visitors now enter the gallery through controlled transition areas as part of the museum’s management of its environment. Such precautions make sense for metal that has spent centuries in seawater and has already undergone several rounds of treatment. Chloride contamination, fragile inlays and the interaction between past restoration and present conditions all require continued attention. Removal from the seabed was never going to be the end of preservation.

What Their Survival Lets Us See

The Riace Bronzes let us see decisions made by Greek sculptors almost at human scale: a hip shifted over the supporting leg, the precise tension of an arm, individual beard curls and eyes assembled to meet a viewer’s gaze. Core materials provide evidence about casting and a broad fifth-century BCE date; the outer surfaces preserve traces of changing marine environments. Much can now be studied directly on the works rather than inferred from Roman copies.

The figures have nevertheless suffered substantial losses. Their weapons have gone, some components were altered in antiquity, and corrosion has changed their surfaces. What remains is an unusually complete record of Classical Greek bronze sculpture, sustained first by escape from recycling and later by careful conservation. The largest gap in their history lies between the unidentified monument for which they were made and the place where Mariottini saw them in 1972. How they travelled between those points is a question their surviving metal may yet help answer.